National EMSC Data Analysis Resource Center
A large part of becoming familiar with your data involves validating the variables in your dataset.
If you collect your own new data, despite all of your best intentions to prevent data errors, somehow data will often still have mistakes in it. Therefore, you will need to validate your data. This means you'll be checking for bad data.
It is important that data validation occur before doing any analyses...If you are using data from an outside source, such as the CDC, you should still validate the data once you have it in the format you plan to analyze it in. This is because there are a lot of unintentional mistakes that can occur when importing or exporting data. Therefore, it is important data validation occurs before doing any analyses or creating tables or reports from the data you plan to use, whether it is new data or data from an outside source.
Before we discuss some important data validation steps, We remind you that throughout the data validation process, you should always be asking yourself the question, "Does this result make sense?"
We will specifically look at: